The Agent Skills Directory

[SAFE]: The skill performs read-only security audits using standard diagnostic tools such as grep, find, and ls. It integrates with well-known security scanners like semgrep, bandit, and pip-audit if available in the environment. All network operations (e.g., npm audit, pip-audit) target official package registries and well-known services. The logic includes specific rules to exclude sensitive directories like .git and node_modules from scans, and it provides clear guidance for severity classification based on industry-standard OWASP categories. No malicious patterns, data exfiltration, or unauthorized persistence mechanisms were detected.

security-review